Pergamon Person. indicid. DijjJ Vol. 21, No. 5, 675486, 1996 pp. Copyright 0 1996 Elsevier Science Ltd Printed in Great Britain. All rights reserved SO191-8869(96)00120-l 0191-8869/96 $15.00+0.00 AGE-RELATED IMPAIRMENT IN ASSOCIATIVE LEARNING: THE ROLE OF ANXIETY, AROUSAL AND LEARNING SELF- EFFICACY John E. Fisk* and Peter Warr Department of Psychology, University of Sheffield, Sheffield SlO 2TN, England zyxwvutsrqponmlkjihgfedcba (Received 7 March 1996) Summary-The impact of individual differences in state anxiety, arousal and learning self-efficacy on performance in an associative learning task was examined. Older people reported higher levels of arousal but lower levels of self-efficacy than younger ones, and no age-different was present in anxiety. Arousal level was negatively related and self-efficacy positively related to learning performance, while anxiety exerted no significant impact. Age differences in learning performance were attenuated following control for arousal and some of this attenuation remained after prior control for age-related differences in perceptual speed. It was concluded that the greater arousal of older participants impaired their performance independently of differences in perceptual speed. Copyright 0 1996 Elsevier Science Ltd. INTRODUCTION In cognitive tasks of at least moderate complexity older people as a group typically perform less well than younger people as a group (e.g. Campbell & Charness, 1990; Fisk & Warr, 1996; Hasher, Stoltzfus, Zacks & Rypma, 1991; Strayer & Kramer, 1994). Most research has examined merely average age-effects, but research attention is now shifting to ask also about differences between individuals. Investigations to identify within-person variables that might underlie average age-patterns have to date been of two principal kinds. Some studies have examined physiological and health-related variations that might account for cognitive differences at different ages. Others have explored aspects of working memory, perceptual speed and related forms of information-processing that might differ between people at different ages. Studies of the former type have shown that differences in physiological health can indeed account for some of the age-difference in cognitive performance (e.g. Houx, Vreeling & Jolles, 1991; Offen- bath, Chodzko-Zajko & Ringel, 1990). Second, it is known that older people process information more slowly than younger ones (e.g. Salthouse, 1985), and this age-related reduction in perceptual speed has also been shown to be implicated at the individual level in a wide range of cognitive deficits (e.g. Salthouse, 1992, 1993; Salthouse & Babcock, 1991). That pattern has been demonstrated specifically in respect of learning by Salthouse (1994) and Fisk and Warr (submitted). However, there have been very few investigations into the role of individual persons’ affective processes in age-differences in cognitive performance. This study seeks to remedy that limitation by addressing two questions. First, are affective features (of anxiety, alertness and learning self-efficacy) important in accounting for the age-pattern observed in an associative learning task? Second, if they are found to be implicated, do these affective features remain significant over and above the established impact on learning of differences in perceptual speed? In respect of the first feature of interest here, several researchers have shown that state anxiety interacts with task complexity to produce performance decrements (e.g. Calvo, Ramos & Estevez, 1992; Hedl & Bartlett, 1989) Belbin and Belbin (1972) have pointed out that older persons are frequently more anxious relative to their younger counterparts when acquiring a new skill. In the context of age differences Charness, Schumann and Boritz (1992) found that while there was no *To whom all correspondence should be addressed. 675